Audyssey problem with channel levels during calibration

[JonnyFive23517]

Before running Audyssey XT setup, I turned Audyssey off, set the Marantz SR6006 to 0 db offset, and adjusted the channel level of each speaker (inc subwoofer) to 75 db at the central listening position. I’ve read this is important to get consistent results.

I ran Audyssey, and the four surround/height speakers were chirping at about 45-47 db at the central listening position, the front L/R and subwoofer about 52-57 db, much lower than the 75 db I had calibrated them to.

At the second listening position, something strange happened. As the front L speaker had just finished its series of chirps (or pings, whatever you call them), I made a loud “creek” by stepping on my wooden stairs. The mic picked up that noise, and the front L speaker re-performed its chirps at a much louder volume, approx 65 db at the main listening position. The chirp volume of all other speakers remained the same.

The image attached shows the end filter result of the front L/R speakers. You can see the L speaker curve is very similar to the R speaker curve, but larger offsets are applied to the L curve, particularly in the high frequency roll off. (Other times I have ran Audyssey the L/R curves are nearly identical, so this isn’t a room acoustics issue).

Question: It’s obvious that the chirp volume level plays an important role in the resulting curve. Why did my 75 db calibration have no effect on chirp volume level?

Question: What happened at the second listening position that caused the L speaker to re-perform its chirps at a higher volume?

Question: What level is the chirp supposed to be??

[Steve81]

Question: It’s obvious that the chirp volume level plays an important role in the resulting curve. Why did my 75 db calibration have no effect on chirp volume level?

Trim levels don't impact the "chirp" level as far as I'm aware.

Question: What happened at the second listening position that caused the L speaker to re-perform its chirps at a higher volume?

Background noise interfering with the measurement. Increasing the volume of the chirp helps to ameliorate this as it improves the signal to noise ratio.

Question: What level is the chirp supposed to be??

Not something I'd spend a lot of time worrying about. Chirp level depends on speaker sensitivity, distance, etc.

[JonnyFive23517]

Background noise interfering with the measurement. Increasing the volume of the chirp helps to ameliorate this as it improves the signal to noise ratio.

So then the question is, why doesn't Audyssey run at a higher chirp volume all the time? Wouldn't it give more accurate results?

[Steve81]

So then the question is, why doesn't Audyssey run at a higher chirp volume all the time?

Might get annoyingly loud in some setups I suppose.

Wouldn't it give more accurate results?

Not materially if the SNR is adequate.

[jostenmeat]

So then the question is, why doesn't Audyssey run at a higher chirp volume all the time? Wouldn't it give more accurate results?

Why don't you ask Audyssey instead? Do you think we would know better than the engineers who created it?

I'm not trying to be snarky; in fact, you may even have limited success by asking them directly, because there is quite a bit of info that they don't share publicly, due to the protection of intellectual property. I mean they won't even share the resolution of their distance measurement! (But Chris K. did mention it was at least finer than 0.1' or something like that.)

Now if you put a gun to my head and forced me to come up with some imagined guesses out of my rear end: maybe they have had more accurate calibrations with the lower level as far as the typical SPL of the reflections/overhang the mics receive. Come to think of it, maybe they calibrated with that lower level with cheaper mics and some inherent limitations in mind, to keep overall added costs down with the final receiver price. If you want some more BS guesses, ask away, and maybe I can come up with more.

If and when you do receiver answers from Audyssey, we would be curious to know as well, so please share!

[JonnyFive23517]

I've got a question in to them, I thought to post here just to see if there were any ideas. I'll let you know what Audyssey says.

[avengineer]

Why don't you ask Audyssey instead? Do you think we would know better than the engineers who created it?

I'm not trying to be snarky;

Well, glad to hear you're not trying to be snarky! I wonder what that would be like.

Now if you put a gun to my head and forced me to come up with some imagined guesses out of my rear end: ... If you want some more BS guesses, ask away, and maybe I can come up with more.

I'll bet the OP didn't want imagined guesses out of any orifice, and I'll bet the thread wouldn't benefit from any more of them either.

The initial answers are basically correct. Audyssey resets any manual cal levels, so they don't matter at all. The chips are done at a lower level first because they are annoying, and if the S/N ratio of the space is good enough, no need for loud chirps. The idea that there could be more accurate measurements at lower levels makes no sense. But, background noise can get in the way, so any amount of background noise during the test that impacts the measurement, and you'll get a second pass at a higher level. If you caused the noise, you can always cancel the calibration and start over, being careful not to squeak the floor this time.

Chirp level is no indication of how the system will play a standard level test track. But, within the Audyssey calibration is level set that compensates for distance, efficiency, placement, etc., and sets system gain so the volume control numbers make sense, i.e., "0" is "Reference Level". Once it's done, you can verify the level with a standard test disc and meter, and find they will reasonably match each other, and hit the right reference levels correctly.